This study seeks to improve estimates of mortality for the African-American population over the period from 1930 to 1990. A central focus is an assessment of the quality of age reporting for black decedents in the 1980s. This assessment will involve direct matching of death certificates for persons reported as dying above age 60 against records for these same individuals in the U.S. Censuses of 1900, 1910, and 1920, and against records of the Social Security Administration. For a sample of persons born in Maryland, the study will also match death certificates against their birth certificates. Based upon these matching studies, corrected distributions of ages at death over the period 1930-1990 will be produced. Using a variety of demographic estimation techniques, data from the matching studies and other data will be converted into improved estimates of black mortality. The methods include intercensal comparisons of cohort size; extinct generation methods; a procedure for producing life tables directly from a distribution of ages at deaths and age-specific growth rates; and a comprehensive method that integrates multiple observations on cohort size from censuses and vital registration into a final estimate of mortality and population. A by-product of this latter method will be new estimates of census completeness. Cause-of-death life tables will also be produced for the African-American population during the 1980-1990 period.